Dataset for LLM-Driven Context-Aware Health Monitoring for Resource-Constrained Edge Devices (Do-Or-Die Framework)

This dataset contains the multi-dimensional system metrics used to evaluate the Do-Or-Die (LLM-Driven, context-aware health mOni-toRing framework for resource-constrained eDge DevIcEs) framework. The data includes real-time time-series representations of system metrics collected from resource-constrained Raspberry Pi edge devices, covering:- CPU usage- Memory utilization- Device temperature- Network load- Process information This data supports the evaluation of the adaptive machine learning model selection orchestrated by a Large Language Model (LLM) layer to balance accuracy, latency, and computational costs under dynamic workloads.